Functional and structural neuroplasticity of somatosensory system in hemiplegic cerebral palsy

Abstract Children with hemiplegic cerebral palsy exhibit diverse sensory deficits linked to varied atypical brain reorganization patterns. The underlying mechanisms of these patterns remain poorly understood. Using multimodal neuroimaging, we associate functional and structural neuroplasticity in the somatosensory system of children with hemiplegic cerebral palsy with lesion type and behavioral outcome. We hypothesize that different lesion types produce distinct patterns of neuroplasticity, which correlate with varying levels of sensory and motor behavioral deficits. We examined 22 children with hemiplegic cerebral palsy (seven with cortical-subcortical, 12 with periventricular, and three with other lesions), and 24 age-matched neurotypical controls. We assessed the somatosensory outcomes through touch sensitivity and two-point discrimination measures, and motor outcomes through movement range, accuracy, dexterity, and fluency measures. We quantified the integrity of ascending sensory and somatosensory commissural fibers with diffusion-weighted MRI and measured cortical responses to haptic stimulation with magnetoencephalography and high-density EEG. Children with cortical-subcortical lesions demonstrated higher sensory thresholds and worse movement quality in the paretic hand compared to the periventricular lesion (P < 0.05) and control group (P < 0.01). Functional imaging revealed suppressed primary and secondary somatosensory cortical activations (P < 0.01) and prolonged primary somatosensory cortex latency (P < 0.05) in the more affected hemisphere in response to paretic hand stimulation. These children also exhibited more pronounced disruption of interhemispheric functional balance (P < 0.05). Tractography showed greater microstructural damage—increased mean (P < 0.01), axial (P < 0.01), and radial (P < 0.01) diffusivities— in both ascending sensory and commissural fibers in the more affected hemisphere of the cortical-subcortical lesion group. In contrast, the periventricular lesion group showed unilateral damage to ascending sensory fibers (P < 0.01) and bilateral damage to commissural fibers (P < 0.05). More severe diffusion abnormalities were associated with lower cortical amplitudes (Rho = –0.58 to – 0.78, P < 0.05), delayed latencies (Rho = 0.69, P < 0.05), and decreased lateralization (Rho= – 0.62 to –0.72, P < 0.05). Moderate to strong correlations were observed between cortical amplitudes and touch sensitivity (Rho = –0.56 to –0.60, P < 0.05) and discriminability (Rho = – 0.77, P < 0.01). Our findings show that children with hemiplegic cerebral palsy exhibit distinct structural and functional neuroplasticity patterns based on lesion type. Cortical-subcortical lesions lead to more extensive somatosensory damage, while periventricular lesions are linked to greater neuroplastic potential. These insights deepen our understanding of neuroplasticity mechanisms and highlight targets for optimizing therapeutic interventions.